Combined low bit rate speech coding and channel coding over a Rayleigh fading channel

In this paper, we present a new soft input speech decoding algorithm in low bit rate speech transmission over a Rayleigh fading channel based on a vector quantization (VQ) technique. Traditionally, in a wireless communication system, if the source coder is designed by VQ based on a clear channel, the technique is called source optimized vector quantization (SOVQ). If the coder is designed based on an optimized channel with a given channel signal-to-noise ratio (SNR), it is called channel optimized vector quantization (COVQ). It has been proven that, based on a first-order Gauss-Markov source, the COVQ soft input decoding algorithm performs much better than the SOVQ and COVQ by using a hard decision decoding over a Rayleigh fading channel. Here, we apply the proposed algorithm directly to a new designed 4.0 kbps VQ-CELP speech coder to improve the decoded speech quality over a Rayleigh fading channel.